Based on the previous results and correlation analyses, multiple regression analyses were made in an attempt to develop statistical "models" that would combine various site factors (soils and management) to evaluate multiple effects on growth and foliage characteristics of trees. As with correlation analyses, these equations assume linear relationships and do not imply direct cause and effect. A summary of those "models" is given in Table 3.
Of all the site factors measured/evaluated, only six were statistically significant in one more of the individual equations, with depth to mottling in the soil, slope shape, and sand in the A horizon occurring in four, thickness of the Ap soil horizon and slope percent in two, and sand in the B horizon in one. In all equations, coefficients for individual site factors were positive, indicating that survival, growth, and/or foliage characteristics of trees improved as slope shape went from concave to convex and depth to mottling, thickness of the Ap soil horizon, sand in the A and B horizons, and slope percent increased (Table 3). All of these relationships are similar to those noted for simple correlations between site factors and individual tree characteristics.
None of the equations accounted for a high percentage of the total variation in tree characteristics, with only 29 percent for total height, 26 percent for four-year height, 20 percent for survival, and 17, 12, and 19 percent, respectively, for needle length, needle color, and foliage weight. Given these small percentages of variation accounted for, it would be meaningless to use the equations to develop guidelines for predicting tree performance on sites having a range of soil and/or topographic conditions.
| Table 3. Summary of Multiple Regression Analyses of Relationships Between Site Factors and Survival, Growth, and Foliage Characteristics of Trees. | ||||||
|---|---|---|---|---|---|---|
| Site Factor | Tree Characteristic | |||||
| Plot Survival2 | Total Height2 | 4-Yr. Height2 | Needle Color2 |
Needle Length2 | Foliage Weight2 | |
| Slope % | + | + | ||||
| Slope Shape | + | + | + | + | ||
| Depth Mtl. | + | + | + | + | ||
| Thick. Ap Horiz. | + | + | ||||
| Sand A Hor. | + | + | + | + | ||
| Sand B Hor. | + | |||||
| Total Variation Acc. For, % | 20 | 29 | 26 | 17 | 12 | 19 |
| 1 Only those site factors that were statistically significant in multiple regression equations were included in the table. 2 A + or a - sign indicates positive or negative relationships between individual site factors and survival or tree characteristics in the multiple regression equation. 3 Percent of total variation (100%) accounted for by one to four factors in individual equations. | ||||||